Skip to content

Code for the paper on t-SNE with variable degree of freedom

Notifications You must be signed in to change notification settings

dkobak/finer-tsne

Repository files navigation

Heavy-tailed kernels reveal a finer cluster structure in t-SNE visualisations

Dmitry Kobak, George Linderman, Stefan Steinerberger, Yuval Kluger, Philipp Berens

https://link.springer.com/chapter/10.1007/978-3-030-46150-8_8

MNIST t-SNE with different degrees of freedom

This is a companion repository to our ECML-2019 paper. All code for the paper is in a Python Jupyter notebook heavy-tailed-tsne.ipynb. We use this t-SNE implementation: https://github.com/KlugerLab/FIt-SNE, version 1.1.0.

For any technical questions, please start an Issue.

                                                      MNIST animation

About

Code for the paper on t-SNE with variable degree of freedom

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages